Research · Human × Agent intelligence
AI made engineers faster. What is the work asking of them now?
Working with AI agents changes the shape of engineering work: more steering and review, more sessions in flight, more switching between them. These briefs set out the questions we're working on, in plain terms.
What we're looking at
Ascenda measures the work, the human doing it, and the relationship between them. These are the areas the briefs draw on.
Supervisory engineering
How much of the week is hands-on work, and how much is steering, reviewing and correcting agents.
Verification and review burden
The trust decisions made on generated output, most of which never show up in review tooling.
Concurrency and context switching
Several agent sessions in flight at once, and the switching between projects and contexts that comes with them.
Work demand
How much the work asked: session shape, interventions, failures and rework, after-hours share.
The human side
What people tell us about how the work landed, in their own words and on their own terms.
Declared changes
Deciding the change, the measure and the window before acting, then reporting the readout whichever way it goes.